Library › DNA & Proteins › Heart muscle — molecules

Heart muscle — molecules: 23 3D models

DNA, RNA, nucleic-acid drugs, gene editing, proteins and 300+ real PDB structures (real sizes 1–100 nm). Each one is a ready-made 3D model you can turn, recolour and place in a figure — open BioDiorama and type its name in the Library search.

Open the 3D Studio

What a failing heart burns — The heart turns over about six kilograms of ATP a day and is never more than a few seconds ahead of demand, so what it burns matters. Normally two thirds comes from long-chain fatty acids, which have to pass CPT1 on the outer mitochondrial membrane. A failing heart turns that gate down without taking up enough glucose to make up the difference — and the one fuel it gets better at using is ketone, which yields more ATP per unit of oxygen. The shortfall shows as a fallen phosphocreatine-to-ATP ratio, which predicts outcome better than the ejection fraction does. (ATP turnover ≈ 6 kg/day)
The cardiac sarcomere — One sarcomere, Z-disc to Z-disc, with everything that makes a heartbeat in it: a thick filament whose myosin heads come off in threes every 14.3 nanometres with a bare zone in the middle, thin filaments reaching in hexagonally from each Z-disc with tropomyosin in the groove and troponin every 38.5 nanometres, the M-line keeping the thick filaments in register, and titin running from Z-disc to M-line. Set the length to see why force peaks where the overlap is best. (1.6–2.3 µm long)
The cross-bridge cycle — Four myosin heads on one thin filament, each caught at a different point of the cycle: detached with ATP bound, cocked once the ATP is hydrolysed, in the power stroke as the phosphate leaves, and in rigor waiting for the next ATP. The lever arm swings about seventy degrees and moves the filament some ten nanometres — and because the heads are out of step, the muscle pulls smoothly rather than in jerks. (stroke ≈ 10 nm)
The troponin switch — What calcium actually does. Troponin C is a dumb-bell of helices with calcium sites at one end; when they fill, troponin I is pulled off the actin and tropomyosin rolls about twenty-five degrees deeper into the groove, uncovering the myosin-binding sites. Take the calcium away and troponin I goes back and holds the filament shut. Switch between diastole and systole to watch the sites open — and see where PKA puts the phosphate that lets calcium come off sooner. (troponin complex ≈ 80 kDa)
The super-relaxed state, and what breaks it — Most myosin heads in a resting heart are not merely idle — they are folded back onto the backbone in pairs, holding each other shut, and using almost no ATP. Myosin-binding protein C is what keeps them there, with its eleven domains lying along the filament and its N-terminus reaching in. A truncating variant removes the end that does the holding, the heads are released, and the muscle burns energy it does not need. Mavacamten folds them back. (thick filament ≈ 1.6 µm)
Titin, the spring in the muscle — Where the Frank–Starling relation comes from. Titin's I-band section is a run of immunoglobulin domains and then the disordered PEVK region; stretch the sarcomere and the PEVK takes up the slack first, and only then do the Ig domains pop open one at a time. That passive tension is what tells the heart how full it is. A truncating variant leaves half a spring, the ventricle gives too easily and dilates — the commonest genetic cause of dilated cardiomyopathy. (the largest human protein, ≈ 3.7 MDa)
The dyad: where a heartbeat begins — A junction twelve nanometres wide. L-type channels in the T-tubule membrane face a cluster of ryanodine receptors on the sarcoplasmic reticulum, and the small trickle of calcium through one channel is enough to open the receptors opposite — calcium-induced calcium release. Everything about the strength of a beat comes down to how well these two membranes stay the right distance apart, which is exactly what fails in heart failure. (gap ≈ 12 nm; RyR2 27 nm across)
RyR2, shut and leaking — The biggest ion channel there is: four subunits making a square foot twenty-seven nanometres across in the cytosol, with a small pore through the SR membrane. Shut, it is held by calstabin-2 and the calcium behind it is held by calsequestrin. Phosphorylated by both PKA and CaMKII and with calstabin come off, it leaks between beats — the calcium goes out by exchange, each ion carrying a net charge inwards, and that small depolarisation can start a beat of its own. (2.2 MDa; foot ≈ 27 nm)
SERCA and its brake — The pump that ends a heartbeat: ten membrane helices and three cytosolic domains that move like a lever, spending one ATP to put two calcium ions back into the store. Phospholamban — a single helix — sits in a groove between its membrane helices and slows it down. PKA puts a phosphate on serine 16, the helix lifts out, the pump runs at full speed and the heart relaxes faster. That is most of what adrenaline does to filling. (SERCA2a ≈ 110 kDa)
The beta-1 receptor and the beat it makes — The pathway a beta blocker interrupts, drawn end to end on one membrane: noradrenaline on the receptor, Gs switched on, adenylyl cyclase making cyclic AMP, PKA, and the three things PKA phosphorylates — the L-type channel, RyR2 on the store and phospholamban. Faster in, more out, and put back sooner: a harder beat that also relaxes quicker. Put a blocker in the pocket and the whole line goes quiet. (receptor ≈ 51 kDa)
BNP, and the enzyme that eats it — The heart's own diuretic hormone and why there are two numbers on the blood test. Corin cuts proBNP in two: BNP, a seventeen-residue ring closed by one disulfide, which does the work, and NT-proBNP, which does nothing but lasts long enough to measure well. Neprilysin then chews BNP up within minutes. Blocking neprilysin lets the hormone survive — which is the second half of what sacubitril-valsartan does. (BNP 32 residues)
ACE, and the inhibitor on its zinc — Angiotensin-converting enzyme has two homologous domains, each with a zinc at the bottom of a deep channel that takes a peptide end-on. Angiotensin I threads in, two residues come off, and what leaves is angiotensin II — eight residues, and the most powerful constrictor the body makes. An ACE inhibitor puts its own zinc-binding group on the metal and nothing is cut. It is not selective: bradykinin builds up too, which is where the cough comes from. (ACE ≈ 170 kDa)
The mineralocorticoid receptor — A nuclear receptor, so it works nothing like a surface one: aldosterone is buried inside the ligand-binding domain, helix 12 closes over it and makes the groove a coactivator docks into, and the two zinc fingers of the dimer grip a response element on the DNA. Sodium is kept, potassium is lost and collagen is laid down. An antagonist pushes helix 12 out of place, no coactivator docks, and the genes stay off. (receptor ≈ 107 kDa)
eNOS: coupled, and uncoupled — Where a healthy blood vessel gets its tone from. In the dimple of a caveola, eNOS makes nitric oxide, which crosses into the smooth-muscle cell, switches on soluble guanylate cyclase, and the muscle relaxes. Short of tetrahydrobiopterin at the dimer interface the enzyme uncouples and makes superoxide from the same oxygen instead — which then destroys the nitric oxide that is left. One enzyme, two opposite jobs, and a cofactor deciding which. (eNOS ≈ 133 kDa per subunit)
Nav1.5 and its inactivation gate — The channel the upstroke of a heartbeat comes through: four domains round one pore, each with six helices, the voltage sensors set outwards and a DEKA filter at the top. What makes it stop is a short linker between domains three and four that swings in and plugs the pore from the inside within a millisecond. Variants that stop it plugging properly leave a late current and lengthen the QT; variants that reduce the current give the Brugada pattern. (≈ 227 kDa)
hERG, and why so many drugs block it — The channel that ends the plateau of a cardiac action potential, and the reason drug development has a cardiac safety assay. Its inner cavity is unusually roomy and lined by a tyrosine and a phenylalanine from each of the four subunits, so molecules with nothing else in common find a snug fit there. Block it and repolarisation is delayed, the QT interval lengthens, and the heart becomes liable to a particular and dangerous arrhythmia. (≈ 127 kDa per subunit)
The funny current, and the drug that slows it — A channel that opens on going negative instead of positive, and passes sodium as well as potassium — so a sinoatrial cell has no resting potential at all. It drifts upward until it fires, and how fast it drifts is set by cyclic AMP binding the domains below the pore, which is how adrenaline speeds the heart without touching anything else. Ivabradine blocks the pore from the inside and slows the rate alone, leaving contractility and blood pressure be. (HCN4 ≈ 129 kDa per subunit)
Gap junctions, and what happens when they move — Heart muscle conducts without synapses because the cells are wired together: hexameric connexons in one membrane dock end to end with those in the next, and the two cytoplasms are continuous. In a healthy ventricle nearly all of them sit at the ends of the cell, so current goes the long way and the wavefront is fast and orderly. In disease they move to the sides — and conduction slows and turns anisotropic, which is how scar makes arrhythmia. (connexon ≈ 9 nm; plaque up to 1 µm)
The cardiac desmosome — What holds one heart cell to the next against a lifetime of pulling: desmoglein-2 and desmocollin-2 shaking hands across a thirty-nanometre gap, an inner plaque of plakoglobin and plakophilin-2, and desmoplakin reaching back to the desmin filaments that take the load. Lose a copy of plakophilin-2 — the gene most often at fault in arrhythmogenic cardiomyopathy — and every beat pulls the cells a little further apart, until fat and fibrous tissue take the muscle's place. (gap ≈ 30 nm)
Transthyretin coming apart — Cardiac amyloidosis starts with a protein that is perfectly stable until it is not. Transthyretin circulates as a tetramer with a thyroxine channel down the middle; it has to dissociate before a monomer can misfold, so that step is the one everything depends on — and the one a stabiliser in the channel prevents. Once free, the misfolded monomers stack in register about five angstroms apart and the fibril grows from both ends. (tetramer 55 kDa; fibril ≈ 10 nm wide)
An amyloid light chain — The other cardiac amyloid, and a different problem entirely: not a normal protein that destabilises with age but an abnormal one made in quantity by a single plasma-cell clone. Two immunoglobulin domains, and the variable one carries that clone's own somatic mutations in its loops — some of which leave it just unstable enough to open, be clipped by a protease, and stack. It is why the treatment is aimed at the clone and not at the heart. (light chain ≈ 23 kDa)
The enzyme that cuts the fibrous cap — A plaque ruptures where its collagen has been cut. MMP-9 holds onto a collagen fibre with its fibronectin modules and puts its catalytic zinc on the triple helix, and one cut at the right place is enough — the rest unwinds and other proteases finish it. Macrophages in the plaque's shoulder make the enzyme; TIMP is the natural wedge that stops it. Whether the cap holds is the balance between those two. (collagen fibril ≈ 1.5 nm per helix)
Sugar stiffening the artery wall — Why arteries get stiff with age and faster with diabetes. Glucose sticks to the lysines of collagen and then rearranges, slowly, into cross-links that no enzyme can undo — so the wall stops giving. The same products are read as injury by RAGE on the cells nearby, which adds inflammation to stiffness. Clinically it shows as a widening pulse pressure: the systolic rises, the diastolic falls, and the heart works against a wall that will not yield. (collagen half-life in artery ≈ 100 days)